Transverse Impact Properties

[lewtam]

Just a purely academic question, as this happened with an employer from which I have since moved on.

Was part of a team that had to update a steel specification for numerous plant items (gears, heavy shafts etc) for a mining equipment manufacturer - to simplify the project, we replaced 40 year old H&T steel specs with slightly more modern case hardening materials. The arguments of H&T vs case hardening steels can be kept for another day - I could start a new forum if the in house arguements is anything to go by.

The members of the team that had the dominant hand were the 3 materials consultants bought in for the job. Of particular concern to them was the transverse impact properties of any steel to be a specified minimum. This was based on, to quote them 'the principles of fracture mechanics'. As one of the site mechanical engineers, this piqued my interest and I asked for a more detailed explanation of how increased impact properties relate to the reduction of crack propagation. Their answer was vague to say the least. Since then, I have tried in vain to locate a source of information that links the simple fracture mechanics that I learned as an undergrad - Kic, cycles to failure etc to impact properties.


I've seen some excellent answers on this forum and I hope someone can help out here.

Thanks in advance.

Lou

 

[MOB1]

This question has arisen many times in material specification. After some research it seems that the answer is:
Crack growth rates in fatigue are in no way related to the critical stress intensity factor or the fracture toughness of the material except that da/dN becomes infinite and total failure occurs during the cycle in which the stress intensity reaches K1c. That is an overload fracture occurs. Therefore the relationship means that the size to which the crack grows before overload fracture occurs is dependent on K1c but not the crack growth rate.
I would refer you to:
ASM Metals Handbook 10th Edition Volume 19 Fatigue & Fracture and to:
Elementary Engineering Fracture Mechanics by David Broek.

 

[Carburize]

There is no direct relationship between true fracture toughness values and Charpy impact test results. Generally the higher the absorbed energy in the Charpy test the higher the fracture toughness. Since Charpy tests are so cheap compared to fracture toughness testing (in the order of 10:1) and there is so much historical data many people have worked on relationships for example Sailors and Corten, Marandet and Sanz, Wells etc etc but all the relationships have been for very narrow groups of steels or very specific conditions.

 

[metengr]

To add to M0B1's post above, there is an indirect correlation between Charpy impact (CVN) and fracture toughness - the critical stress intensity factor (KIc) for certain steels.

My guess is that the materials consultants were concerned with anisotropic fracture toughness of the gear components in service. The critical stress intensity factor is effected by grain orientation (aka anisotropic behavior) and was probably the reason for using minimum transverse impact values as a screening method. As the toughness of the material is reduced from microstructural anisotropy, the tolerance to flaws is also reduced rendering the part susceptible to fracture in service.

There have been several empirical correlations developed to estimate the equivalent KIc value from Charpy impact tests. The two common methods are Barsom and Rolfe, and Sailors and Corten (see Fracture Mechanics of Engineering Materials by Hertzberg).

 

[lewtam]

Thanks guys! A loose empirical relationship is enough for me. I have access to a copy of Herzberg (local university library) so will be able to read into it.

Metengr - the comment on grain orientation and KIc is interesting and I will make sure I familiarize myself with this.

Just for info - new spec ended up as near to a copy of AGMA C95 as possible without them having to attribute the info to AGMA, ie lower maximum allowable sulfur a bit more to make it different. The stipulations are in the same order as they occur in C95.

The new spec has been in service for about 2 years now - no failures attributed to material quality.

Now if someone has established an empirical relationship between customer's expectations and the incidence of low speed overload...